The role of religious advisors in mental health care in the World Mental Health surveys

Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 DOI 10.1007/s00127-016-1290-8 ORIGINAL PAPER The role of religious advisors in mental health care in the World Mental Health surveys Vivianne Kovess-Masfety 1 Sara Evans-Lacko 2 David Williams 3 Laura Helena Andrade 4 Corina Benjet 5 Margreet Ten Have 6 Klaas Wardenaar 7 Elie G. Karam 8 Ronny Bruffaerts 9 Jibril Abdumalik 10 Josep Maria Haro Abad 11 Silvia Florescu 12 Benjamin Wu 13 Peter De Jonge 14 Yasmina Altwaijri 15 Hristo Hinkov 16 Norito Kawakami 17 Jose Miguel Caldas-de-Almeida 18 Evelyn Bromet 19 Giovanni de Girolamo 20 José Posada-Villa 21 Ali Al-Hamzawi 22 Yueqin Huang 23 Chiyi Hu 24 Maria Carmen Viana 25 John Fayyad 26 Maria Elena Medina-Mora 27 Koen Demyttenaere 28 Jean-Pierre Lepine 29 Samuel Murphy 30 Miguel Xavier 31 Tadashi Takeshima 32 Oye Gureje 10 Received: 26 October 2015 / Accepted: 10 October 2016 / Published online: 2 November 2016 Springer-Verlag Berlin Heidelberg 2016 Abstract Objectives To examine the role of religious advisors in mental health care (MHC) according to disorder severity, socio-demographics, religious involvement and country income groups. Methods Face to face household surveys in ten high income (HI), six upper-middle income (UMI) and five low/ lower-middle (LLMI) income countries totalling 101,258 adults interviewed with the WMH CIDI plus questions on use of care for mental health problems and religiosity. Results 1.1% of participants turned to religious providers for MHC in the past year. Among those using services, 12.3% used religious services; as much as 30% in some LLMI countries, around 20% in some UMI; in the HI income countries USA, Germany, Italy and Japan are between 15 and 10% whenever the remaining countries are & Vivianne Kovess-Masfety vkovess@gmail.com 1 2 3 4 5 6 7 8 EA 4057 Paris Descartes University Paris, UFR Institut de Psychologie, 71, avenue Edouard Vaillant, 92774 Boulogne-Billancourt, France Kings College London, Institute of Psychiatry, Psychology and Neuroscience Health Service and Population Research, London, UK Department of Society, Human Development and Health, Harvard School of Public Health, Boston, USA Institute of Psychiatry, University of Sao Paulo Medical School, Sao Paulo, Brazil Department of Epidemiologic and Psychosocial Research, National Institute of Psychiatry Ramón de la Fuente, Mexico, Mexico Netherlands Institute of Mental Health and Addiction, Epidemiology, Utrecht, Netherlands Department of Psychiatry, University of Groningen, University Medical Center Groningen (UMCG), Interdisciplinary Center Psychopathology and Emotion Regulation (ICPE), Groningen, The Netherlands IDRAAC, Psychiatry and Clinical Psychology, Ashrafieh, Beirut, Lebanon 9 10 11 12 13 14 15 16 17 Katholieke Universiteit Leuven (UPC-KUL), Universitair Psychiatrisch Centrum Leuven, Kortenberg, Belgium Department of Psychiatry, University of Ibadan, Ibadan, Nigeria Departament de Recerca Sant Boi de Llobregat, Parc Sanitari Sant Joan de Déu CIBERSAM, Barcelona, Spain National School of Public Health and Professional Development, Research and Evaluation, Bucharest, Romania Department of Health Care Policy, Harvard Medical School, Boston, USA Department of Developmental Psychology, Faculty of Social and Behavioural Sciences, University of Groningen, University Medical Center Groningen, Interdisciplinary Center Psychopathology and Emotion Regulation, Groningen, The Netherlands King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia Department of Global Mental Health, National Center for Public Health Protection, Sofia, Bulgaria Department of Mental Health, School of Public Health, The University of Tokyo, Tokyo, Japan

354 Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 much lower. In LLMI 20.9% used religious advisors for the most severe mental disorders compared to 12.3 in UMI and 9.5% in HI. For severe cases most of religious providers use occurred together with formal care except in Nigeria, Iraq and Ukraine where, respectively, 41.6, 25.7 and 17.7% of such services are outside any formal care. Frequency of attendance at religious services was a strong predictor of religious provider usage OR 6.5 for those who attended over once a week (p \ 0.0001); as seeking comfort often through religion in case of difficulties OR was 3.6 (p = 0.004) while gender and individual income did not predict use of religious advisors nor did the type of religious affiliation; in contrast young people use them more as well as divorced and widowed OR 1.4 (p = 0.02). Some country differences persisted after controlling for all these factors. Conclusions Religious advisors play an important role in mental health care and require appropriate training and collaboration with formal mental healthcare systems. Religious attitudes are strong predictors of religious advisors usage. Keywords Religion Mental health Services use Background Despite spectacular advances in technology and sciences, 90% of the world s population is involved today in some form of religious and spiritual practice. Non religious people make up less than 1.1% in population in many middle Eastern and African countries. Religion is unlikely to disappear in the world and the role of religious advisors in providing services for mental health problems is likely to remain stable or increase rather than to decrease. adapted from [1]. These sentences apply to many countries at diverse income levels and for some of the highest income countries to their growing immigrant populations. In a prior publication, the World Mental Health Initiative (WMH 1 ) provided data on nationally representative samples of a large set of diverse countries showing a dramatic treatment gap across the world: a considerable portion of individuals with severe mental disorders did not receive any services in the previous 12 months. In addition, who have received services, access to specialty mental health services remained quite low. This trend was, as expected, more visible in low to lower middle income countries. It remained, however, unclear to what extent religious advisors played a role in the use of services because religious advisors were pooled into the human services category along with social workers and counselors [2]. Access to religious providers does not require referral and is free of charge. As such, it could be the only available resource for people with low financial resources [3]. However, there have been some concerns about the quality of care that is delivered by religious providers. Although they are in contact with the most impaired [4], the number of counseling visits together with the low level of cooperation and referral to the formal MH care system have been underlined [3, 5]. Moreover, the mental health training of religious providers, even though some progress has been noticed, seems very poor. Studies have described religious providers inability to identify mental health problems and their underestimation of severity, including 18 19 20 21 22 23 24 25 26 Chronic Diseases Research Center (CEDOC) and Department of Mental Health, Faculdade de Ciencias Medicas Lisbon, Universidade Nova de Lisboa, Lisbon, Portugal Department of Psychiatry, Stony Brook University, Stony Brook, NY, USA IRCCS St. John of God Clinical Research Centre, Brescia, Italy Saldarriaga Concha Foundation, Development Rehabilitation System FSC Bogotá, Bogota, Colombia College of Medicine Diwania Governate, Al-Qadisiyah University, Diwania, Iraq Institute of Mental Health, Peking University, Beijing, China Shenzhen Institute of Mental Health and Shenzhen Kangning Hospital, Shenzhen, Guangdong, China Department of Social Medicine, Federal University of Espírito Santo (UFES), Vitoria, Brazil Institute for Development, Research, Advocacy, and Applied Care (IDRAAC), Beyrut, Lebanon 27 28 29 30 31 32 Instituto Nacional de Psiquiatria, Epidemiology and Psychosocial Research, Mexico, Mexico Department of Psychiatry Leuven, University Hospital Gasthuisberg, Louvain, Belgium Department of Psychiatry, Fernand Widal Hospital, Paris, France Psychology Research Institute Londonderry, University of Ulster, Coleraine, UK Medical Sciences Lisboa, University Nova of Lisboa, Lisbon, Portugal Japan National Institute of Neurology and Psychiatry, Epidemiology, Tokyo, Japan 1 http://www.hcp.med.harvard.edu/wmh/.

Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 355 suicide lethality [3]. Also, some religious providers do not believe in biological models of mental health disorders and attribute the problems to religious causes, lack of spirituality, or usage of drugs or alcohol [6, 7]. Due to these false beliefs, religious providers in some cultures oppose medication [7] and/or consider psychiatry a threat for their interventions [8]. At any rate, a very high percentage of religious providers reported being frequently in contact with people suffering from mental health problems and feeling inadequate to ask for more training of formal MH resources [6, 9]. Most authors quoted above agree that religious providers are the entry door for many people that suffer from mental health disorders and often the only directly available resource for these people. As such, their training and cooperation is essential for delivering adequate care to these patients. To turn to a religious advisor for a mental health problem depends on a variety of factors in addition to religion and religiosity, including: age, gender, social class, education, race and clinical characteristics of the mental health problem [5, 10]. These factors have been relatively well studied in high income countries. However, little is known regarding whether and how these factors apply to lower or middle income countries where religious providers may play a prominent role. The WMH surveys offer a unique opportunity to study the role of religious providers in mental health care in a large sample of countries whose income level, religions, and availability of mental health resources vary enormously. The objectives of this study are (1) to examine the role of religious advisors in MH care together with or without formal care, according to disorder severity; (2) to characterize the sociodemographic and clinical characteristics associated with the use of religious providers for mental health problems and (3) to evaluate the relationship between type of religious affiliation and intensity of religious participation and use of religious providers for mental health care across low, middle and high income countries. Methods Sample The WMH surveys are epidemiological surveys of prevalence and correlates of commonly occurring mental disorders [11] administered in ten countries classified by the World Bank at the time of each survey, as high income (HI) (Belgium, Germany, Italy, Japan, the Netherlands, Northern Ireland, Portugal, Spain, France and the United States), six upper-middle income (UMI) (Brazil, Bulgaria, Lebanon, Mexico, Romania and South Africa), and five low/lower-middle income (LLMI) [Colombia, Iraq, Nigeria, Peoples Republic of China (two surveys one in Beijing/Shanghai and another in Shenzhen) and Ukraine] (World Bank 2009). Most surveys featured nationally representative household samples, while two (Colombia, Mexico) were representative of all urban areas in the country, one of selected states (Nigeria), and three of selected Metropolitan Areas (Brazil, Japan, Peoples Republic of China). A total of 101,258 adults (age 18?) participated. Sample sizes ranged from 2357 (Romania) to 9282 (United States). Informed consent was obtained using procedures approved by local Institutional Review Boards. The average weighted response rate was 73.4% (45.9 95.2% range). Weights were used to adjust for differential probabilities of selection and discrepancies with population socio-demographic/geographic distributions. (see Table 1). Subsampling was used in most surveys to reduce respondent burden by dividing the interview into two parts. Part I, administered to all respondents, assessed core DSM- IV mental disorders (n = 101,258 respondents across all 22 surveys). Part II assessed additional disorders and correlates. Questions regarding service use and religious preference were included in Part II, which was administered to 100% of Part I respondents who met lifetime criteria for any Part I disorder and a probability subsample of other Part I respondents (n = 50,134 across all 22 surveys). Part II respondents with no Part I disorder were upweighted to adjust for under-sampling. Additional weights adjusted for differential within and between household selection and deviations between sample and population demographic-geographic distributions. Further details about WMH sampling and weighting are available elsewhere [12]. Assessment The surveys utilized the lay-administered Composite International Diagnostic Interview (CIDI 3.0; [11]) translated into each local language, to assess lifetime and 12-month diagnoses for all DSM-IV mood, anxiety, substance use and impulse control disorders, and collect information on age of onset, severity, impairment and persistence of disorders. Extensive information is also collected on demographic, behavioural and socioeconomic risk factors. We classified WMH-CIDI mental disorders as serious, moderate, or mild. If any of the following conditions are met the 12-month disorder is considered serious: mania or substance dependence with a physiological dependence syndrome, a suicide attempt in conjunction with

356 Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 Table 1 WMH sample characteristics by World Bank income categories a Country by income category Survey b Sample characteristics c Field dates Age range Sample size Part 1 Part 2 Part 2 B44 d Response rate e I Low and lower-middle Colombia NSMH All urban areas of the country (approximately 73% 2003 18 65 4426 2381 1731 87.7 of the total national population) Iraq IMHS Nationally representative 2006 7 18 96 4332 4332 95.2 Nigeria NSMHW 21 of the 36 states in the country, representing 57% of the national population. The surveys were conducted in Yoruba, Igbo, Hausa and Efik languages 2002 3 18 100 6752 2143 1203 79.3 PRC f Beijing/ Shanghai B-WMH/S- WMH Beijing and Shanghai metropolitan areas 2002 3 18 70 5201 1628 570 74.7 Shenzhen Shenzhen metropolitan area. Included temporary PRC f 2006 7 18 88 7132 2475 80.0 Shenzhen g residents as well as household residents Ukraine g CMDPSD Nationally representative 2002 18 91 4724 1719 540 78.3 TOTAL 32,567 14,678 4044 81.4 II Upper-middle Brazil São Paulo São Paulo Megacity São Paulo metropolitan area 2005 7 18 93 5037 2942 81.3 Bulgaria NSHS Nationally representative 2003 7 18 98 5318 2233 741 72.0 Lebanon Lebanon Nationally representative 2002 3 18 94 2857 1031 602 70.0 Mexico M-NCS All urban areas of the country (approximately 75% 2001 2 18 65 5782 2362 1736 76.6 of the total national population) Romania RMHS Nationally representative 2005 6 18 96 2357 2357 70.9 South SASH Nationally representative 2003 4 18 92 4315 4315 87.1 Africa g TOTAL 25,666 15,240 3079 76.6 III High Belgium ESEMeD Nationally representative. The sample was selected 2001 2 18 95 2419 1043 486 50.6 from a national register of Belgium residents France ESEMeD Nationally representative. The sample was selected from a national list of households with listed telephone numbers 2001 2 18 97 2894 1436 727 45.9 Germany ESEMeD Nationally representative 2002 3 18 95 3555 1323 621 57.8 Italy ESEMeD Nationally representative. The sample was selected 2001 2 18 100 4712 1779 853 71.3 from municipality resident registries Japan WMHJ Eleven metropolitan areas 2002 6 20 98 4129 1682 55.1 2002 2006 Netherlands ESEMeD Nationally representative. The sample was selected 2002 3 18 95 2372 1094 516 56.4 from municipal postal registries N. Ireland NISHS Nationally representative 2004 7 18 97 4340 1986 68.4 Portugal NMHS Nationally representative 2008 9 18 81 3849 2060 1070 57.3 Spain ESEMeD Nationally representative 2001 2 18 98 5473 2121 960 78.6 United NCS-R Nationally representative 2002 3 18 99 9282 5692 3197 70.9 States TOTAL 43,025 20,216 8,430 66.8

Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 357 Table 1 continued Country by income category Survey b Sample characteristics c Field dates Age range Sample size Part 1 Part 2 Part 2 B44 d Response rate e IV Total 101,258 50,134 15,553 73.4 a The World Bank (2008) Data and Statistics. Accessed May 12, 2009. http://go.worldbank.org/d7sn0b8yu0 b NSMH (The Colombian National Study of Mental Health); IMHS (Iraq Mental Health Survey); NSMHW (The Nigerian Survey of Mental Health and Wellbeing); B-WMH (The Beijing World Mental Health Survey); S-WMH (The Shanghai World Mental Health Survey); CMDPSD (Comorbid Mental Disorders during Periods of Social Disruption); NSHS (Bulgaria National Survey of Health and Stress); LEBANON (Lebanese Evaluation of the Burden of Ailments and Needs of the Nation); M-NCS (The Mexico National Comorbidity Survey); RMHS (Romania Mental Health Survey); SASH (South Africa Health Survey); NSMHWB (National Survey of Mental Health and Wellbeing); ESEMeD (The European Study Of The Epidemiology Of Mental Disorders); WMHJ2002-2006 (World Mental Health Japan Survey); NISHS (Northern Ireland Study of Health and Stress); NMHS (Portugal National Mental Health Survey); NCS-R (The US National Comorbidity Survey Replication) c Most WMH surveys are based on stratified multistage clustered area probability household samples in which samples of areas equivalent to counties or municipalities in the US were selected in the first stage followed by one or more subsequent stages of geographic sampling (e.g., towns within counties, blocks within towns, households within blocks) to arrive at a sample of households, in each of which a listing of household members was created and one or two people were selected from this listing to be interviewed. No substitution was allowed when the originally sampled household resident could not be interviewed. These household samples were selected from Census area data in all countries other than France (where telephone directories were used to select households). Several WMH surveys (Belgium, Germany, Italy) used municipal resident registries to select respondents without listing households. The Japanese sample is the only totally un-clustered sample, with households randomly selected in each of the 11 metropolitan areas and one random respondent selected in each sample household. 15 of the 21 surveys are based on nationally representative household samples d Brazil, Iraq, Japan, Northern Ireland, PRC Shenzhen, Romania, and South Africa did not have an age restricted Part 2 sample. All other countries, with the exception of Nigeria, PRC (B-WMH; S-WMH), and Ukraine (which were age restricted to B39) were age restricted to B44 e The response rate is calculated as the ratio of the number of households in which an interview was completed to the number of households originally sampled, excluding from the denominator households known not to be eligible either because of being vacant at the time of initial contact or because the residents were unable to speak the designated languages of the survey. The weighted average response rate is % f People s Republic of China g For the purposes of cross-national comparisons we limit the sample to those 18? any other disorder, reporting severe role impairment due to a mental disorder in at least two areas of functioning measured by disorder-specific Sheehan Disability Scales (SDS; [13]) or having overall functional impairment from any disorder consistent with a Global Assessment of Functioning (GAF; [14]) score of 50 or less. Disorders were classified as moderate if the respondent had substance dependence without a physiological dependence syndrome or at least moderate interference in any SDS domain. All other disorders were classified as mild. Significant monotonic associations have been found in all but two surveys between disorder severity and number of days in the previous year that respondents were unable to undertake normal daily activities because of disorders [15]. All participants were asked if they received any services for problems with emotions, nerves, or your use of alcohol or drugs. Individuals reporting any lifetime use of services were then asked to select whom they had seen from a list of formal health care providers: psychiatrists, psychologists, psychotherapists, social workers, counselors, nurses, general practitioners, other medical doctors. Two categories of informal care providers were defined (1): religious or spiritual advisors and (2) all other nonformal, non-religious providers including the social workers and counselors who are not practicing in a health care setting and complementary alternative medicine including healers and self-help groups. Participants were also asked for their religious affiliation. Among those respondents declaring a religious affiliation, four additional items were asked regarding the nature of their religious practice and intrinsic religiosity including: (1) How often do you usually attend religious services (dichotomized more than once a week vs all others)? (2) How important are religious beliefs in your daily life (dichotomized very important vs all others)? (3) How often do you seek comfort through religious or spiritual means when you have problems or difficulties in your family, work or personal life (often vs all others)? (4) When making decisions in your daily life, how often do you think about what your religious or spiritual beliefs suggest you should do (often vs. all others)? Depending on the analysis, the raw, ordinal versions of the variables or the dichotomized versions were used.

358 Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 Statistical analysis Cross-tabulation was used to examine, among respondents that received services, the proportion that received care in formal settings only, and religious service only, and in the two setting combined. These analyses were carried out in subsamples defined by severity of disorder by each country, by countries combined by income level, and all countries combined. Logistic regression analysis was used to study correlates of declaring a religious affiliation as well as predictors of receiving religious treatment who received treatment in the past 12 months. Standard errors were estimated using the Taylor series method to adjust the weighting as well as for the geographic clustering of observations both between and within countries. These adjustments were implemented using the SUDAAN (version 8.0.1). The coefficients in the logistic regression equations and their design-based standard errors were transformed into odds ratios (ORs) and 95% confidence intervals for ease of interpretation. Multivariate significance tests in the logistic regression analyses were made using Wald v 2 tests based on coefficient variance covariance matrices, adjusted for design effects using the Taylor series method. Statistical significance was evaluated using 2-sided design-based tests and a 0.05 level of significance. Results Use of religious advisor for mental health problems As previously reported [2], the 12-month use of any services for mental health problems largely differs among income group countries ranging from 12.1% in the high income group to 8.7% in the upper-middle to 3.6% in the LLMI with an average of 8.6%. 1.1% of the population declared a religious provider use (together with formal care or exclusively): 0.6% in LLMI, 1.4% in UMI and 1.1% in the HI group. Among those who used services, religious provider usage averaged 12.3% (16.4% in LLMI, 16.7% in UMI to 9.2% in HI countries Table 2). In each income group some country differences were noted that render these groups relatively heterogeneous. The greatest contrast being within the LLMI where usage was very low in China (7.5%) and in the Shenzhen sample (2.9%) and quite high in Iraq, Nigeria and Ukraine (30.3, 30.9 and 24.9%, respectively). In the UMI group, South Africa had a high percentage (24.9%), followed by Lebanon and Brazil while the remaining countries of the group had rather low usage rates. In HI countries, the USA had the highest rate (15.5%), followed by Germany (12.2%), Japan (11.6%), and Italy (9.1%) while the remaining countries had much lower rates. The proportion of respondents receiving care from religious advisors (religious only or combined with formal) varied by disorder severity (Table 2). In LLMI countries, 20.6% of those with a severe disorder had contact with a religious advisor, compared to 12.3% in UMI and 9.5% in HI countries. The contrast is even greater for cases of severe disorders that turn to religious providers as the only resource: 16.2% in LLMI countries versus 3.6% in UMI and 2.5% in HI countries. For instance, in Nigeria, 41.6% of the most severe cases are treated by religious advisors only. Conversely, the use of religious advisors as the only treatment for severe disorders is consistently low in HI and MHI countries. In the WMH surveys, the Catholic religion was most prominent followed by Protestantism and thirdly Islam, however, the distribution varies importantly by country (see Table 3). The propensity to turn to a religious provider did not vary much as a function of religion. Muslims tended to turn to religious advisors slightly more than other groups (24.2%), followed by those declaring other religions (including Judaism, and other non-christian and non- Eastern religions 21.7%), Protestants (18.4%, including both evangelical and non-evangelical groups), and Hindus/ Buddhists (15%). Catholics and those declaring no religious affiliation had the lowest proportion of seeking religious services for a mental health problem (8.8 and 3.3%, respectively) (table available on request). In an adjusted prediction equation, (controlling for socio-demographic characteristics, illness severity, pooled across countries), who had used any services, women and young people were more likely to seek help from religious advisors. Illness severity and personal income were not related to seeking help from a religious advisor (results available upon request). In the fully adjusted logistic equation that added religious preference and questions on religiosity, the effect of gender disappeared (Table 4). Five main findings stand out. First, younger age remained significantly associated with a greater likelihood of using religious advisors for MH care. Second, those separated or widowed had higher usage than those married or never married. Third, religious affiliation was not a predictor of seeking help from a religious advisor while the frequency of attendance at a religious service was a strong predictor of religious provider usage, but only if attendance was higher than once a month. Fourth, seeking comfort through religion in case of difficulties was also a main predictor. Fifth, living in certain countries was associated with using religious providers for mental health problems. The model was done separately for each of the income groups and revealed similar results except for two countries: Iraq with an OR of 0.2 (p = 0.05) and Lebanon with an OR of 2.3 (p = 0.03) (Table 4).

Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 359 Table 2 Service use and severity of disorder who used any services Country Service category Service types w/serious DX severity Service types w/moderate DX severity Service types w/mild DX severity Service types w/no DX severity Services types total % SE % SE % SE % SE % SE China Formal only 76.6 17.1 49.3 25.0 100.0 0.0 74.6 7.2 71.1 8.0 Religious only 0.0 0.0 0.0 0.0 0.0 0.0 5.4 3.4 4.3 2.7 Any religious service 14.7 14.4 0.0 0.0 0.0 0.0 8.9 4.7 7.5 3.8 Colombia Formal only 83.4 8.7 92.3 4.2 95.4 4.5 86.3 4.6 87.5 3.2 Religious only 4.5 2.8 1.5 1.5 4.6 4.5 5.9 3.1 4.7 1.9 Any religious service 12.3 7.3 2.5 1.9 4.6 4.5 8.5 3.8 8.0 2.7 Iraq Formal only 72.4 5.8 39.0 10.8 50.5 9.6 80.0 3.2 65.2 5.0 Religious only 25.7 5.7 55.7 11.4 31.2 3.9 18.0 3.1 29.9 4.3 Any religious service 26.9 5.8 55.7 11.4 31.2 3.9 18.0 3.1 30.3 4.3 Nigeria Formal only 58.4 22.9 88.1 12.3 96.5 4.0 53.5 13.7 68.0 10.1 Religious only 41.6 22.9 11.9 12.3 0.0 0.0 43.7 14.0 29.4 10.1 Any religious service 41.6 22.9 11.9 12.3 3.5 4.0 44.6 13.9 30.9 10.1 Shenzhen Formal only 34.1 17.6 27.9 4.8 58.8 8.5 42.1 5.9 41.8 3.8 Religious only 11.8 11.2 0.0 0.0 3.2 3.0 2.1 1.7 2.0 1.3 Any religious service 26.9 16.5 4.6 3.1 3.2 3.0 2.1 1.7 2.9 1.4 Ukraine Formal only 71.3 7.7 77.5 8.0 65.8 13.3 51.2 11.5 62.5 7.2 Religious only 17.7 6.2 12.8 6.4 24.3 11.5 28.7 11.7 22.4 6.5 Any religious service 20.9 6.4 16.5 6.8 29.7 12.5 30.0 11.6 24.9 6.4 LOW/LOWER MIDDLE-INCOME Formal only 75.3 4.0 63.2 4.2 74.6 4.1 65.2 3.2 67.7 2.2 Religious only 16.2 2.7 16.6 4.4 12.5 2.0 13.4 2.8 14.4 1.8 Any religious service 20.6 3.5 18.5 4.4 13.8 2.2 14.9 2.9 16.4 1.9 Brazil Formal only 69.6 4.4 74.3 8.5 88.5 4.4 78.3 4.3 75.9 2.4 Religious only 2.7 1.4 9.3 6.2 6.6 3.7 6.0 2.0 5.6 1.6 Any religious service 14.5 3.5 15.3 6.9 8.0 3.8 14.9 3.5 13.9 2.4 Bulgaria Formal only 97.4 2.6 99.4 0.7 89.2 10.3 96.5 1.8 95.9 2.3 Religious only 2.6 2.6 0.0 0.0 10.8 10.3 1.7 1.3 2.9 1.8 Any religious service 2.6 2.6 0.0 0.0 10.8 10.3 2.7 1.6 3.5 1.9 Lebanon Formal only 90.5 6.7 87.2 8.0 92.1 8.1 76.6 9.9 82.2 5.9 Religious only 9.5 6.7 12.8 8.0 7.9 8.1 17.9 9.9 14.7 6.2 Any religious service 9.5 6.7 12.8 8.0 7.9 8.1 23.4 9.9 17.8 5.9 Mexico Formal only 85.5 5.3 80.2 7.5 74.2 11.3 71.7 4.8 75.9 3.4 Religious only 1.3 1.3 2.4 2.4 5.0 3.4 3.6 2.9 3.2 1.7 Any religious service 1.3 1.3 3.9 2.6 5.0 3.4 8.8 3.6 6.2 2.0 Romania Formal only 98.4 1.5 89.2 8.6 100.0 0.0 92.5 3.4 94.7 1.8 Religious only 1.6 1.5 10.8 8.6 0.0 0.0 1.2 1.2 2.3 1.1 Any religious service 1.6 1.5 10.8 8.6 0.0 0.0 1.2 1.2 2.3 1.1 S. Africa Formal only 53.3 10.2 55.7 6.2 51.7 8.2 57.9 3.3 56.7 2.8 Religious only 7.1 4.2 3.9 2.2 21.9 6.5 12.9 1.9 12.5.4 Any religious service 21.6 8.3 29.7 5.5 23.5 6.9 24.7 2.4 24.9 1.9 UPPER MID-INCOME Formal only 73.5 3.0 72.6 3.7 73.2 4.0 68.2 2.1 70.3 1.5 Religious only 3.6 1.1 6.0 2.4 12.6 3.3 9.6 1.1 8.4 0.8 Any religious service 12.3 2.5 17.0 3.2 13.6 3.4 18.5 1.6 16.7 1.2

360 Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 Table 2 continued Country Service category Service types w/serious DX severity Service types w/moderate DX severity Service types w/mild DX severity Service types w/no DX severity Services types total % SE % SE % SE % SE % SE Belgium Formal only 90.8 4.9 100.0 0.0 97.6 2.4 85.7 5.4 90.0 3.5 Religious only 0.0 0.0 0.0 0.0 0.0 0.0 2.1 2.2 1.1 1.2 Any religious service 4.4 3.5 0.0 0.0 0.0 0.0 5.0 3.1 3.8 1.8 France Formal only 92.0 4.4 93.1 4.2 88.3 6.5 93.1 3.6 92.3 2.4 Religious only 5.3 3.9 0.0 0.0 9.1 6.0 0.0 0.0 2.1 1.0 Any religious service 8.0 4.4 0.0 0.0 9.1 6.0 1.9 1.1 3.4 1.2 Germany Formal only 88.9 6.0 90.2 3.9 84.5 8.1 76.1 7.0 80.4 4.6 Religious only 0.0 0.0 2.2 2.2 9.4 6.8 12.2 6.5 9.1 4.4 Any religious service 6.3 4.4 3.5 2.5 13.8 7.6 14.9 6.7 12.2 4.5 Italy Formal only 94.6 3.6 92.2 4.0 96.8 3.3 85.2 5.7 90.1 2.6 Religious only 0.0 0.0 6.2 3.7 0.0 0.0 3.9 2.1 3.3 1.6 Any religious service 5.4 3.6 6.2 3.7 0.0 0.0 14.8 5.7 9.1 2.4 Japan Formal only 61.2 13.4 56.7 16.5 86.8 10.9 72.9 6.0 71.1 4.7 Religious only 0.0 0.0 5.7 4.5 1.9 2.0 9.7 6.9 7.7 4.7 Any religious service 0.0 0.0 10.3 6.3 1.9 2.0 14.6 7.2 11.6 5.1 N. Ireland Formal only 80.0 4.7 82.7 6.3 77.3 8.1 92.7 2.5 84.7 2.5 Religious only 0.5 0.5 0.9 0.9 1.0 1.1 3.1 1.6 1.6 0.6 Any religious service 2.3 1.0 3.0 2.2 2.8 2.1 3.9 1.6 3.1 0.8 Netherlands Formal only 85.3 5.7 92.2 3.5 64.0 19.3 81.5 6.0 82.2 4.1 Religious only 0.9 0.9 0.7 0.8 5.2 3.8 0.0 0.0 0.7 0.2 Any religious service 6.8 3.3 0.7 0.8 5.2 3.8 5.3 2.3 5.0 1.5 Portugal Formal only 91.0 2.7 91.0 2.6 86.7 3.5 88.4 3.1 89.4 1.9 Religious only 0.4 0.4 1.5 0.9 0.0 0.0 2.2 0.9 1.5 0.5 Any religious service 4.0 1.8 4.8 1.9 0.9 0.9 4.0 1.2 3.9 0.9 Spain Formal only 96.2 2.3 93.7 3.9 88.0 5.8 96.1 1.3 94.8 1.2 Religious only 0.0 0.0 0.0 0.0 0.0 0.0 0.7 0.7 0.3 0.3 Any religious service 3.0 1.8 0.9 0.7 6.8 5.1 1.5 0.9 2.1 0.8 USA Formal only 66.1 2.4 68.4 2.5 72.8 5.9 69.9 2.6 69.1 1.4 Religious only 4.7 1.0 6.7 1.2 4.8 1.7 9.8 2.1 7.2 1.0 Any religious service 16.4 2.1 15.1 1.8 16.1 3.7 13.9 2.1 15.1 1.2 HIGH-INCOME Formal only 77.4 1.6 79.7 1.6 78.5 3.5 80.6 1.4 79.5 0.9 Religious only 2.5 0.5 3.9 0.6 3.9 1.0 5.7 1.0 4.4 0.5 Any religious service 9.5 1.1 8.7 0.9 10.5 2.2 9.0 1.1 9.2 0.7 ALL COUNTRIES Formal only 76.2 1.4 76.3 1.5 76.5 2.4 74.4 1.1 75.3 0.7 Religious only 4.3 0.5 5.8 0.9 7.5 1.2 8.1 0.8 6.8 0.4 Any religious service 11.4 1.0 11.6 1.1 11.8 1.7 13.0 0.9 12.3 0.6 Religious affiliation and religious involvement The majority of respondents declared a religious affiliation (Table 5), with the exception of the two Chinese samples where less than 10% of the respondents did so. In 11 out of the 22 surveyed countries more than 90% of the respondents declared a religion. Predictors of declaring a religion, in addition to strong country effects, were: women (OR 1.5; 95% CI 1.3 1.6), aged 50 64 (OR 1.4; 95% CI 1.2 1.6) or over 65 (OR 2.3; 95% CI 1.9 2.9), marital status (not married OR 0.8; 95% CI 0.7 0.9), individual income (High income OR 0.8; 95% CI 0.7 0.9),

Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 361 Table 3 Religion type by income group Religion Freq Percent Lower Upper Low low middle income countries No religion 3833 28.1 26.9 29.2 Catholic 3146 22.0 20.5 23.4 Protestant 1846 12.5 11.6 13.4 Hindu/Buddhist 336 1.8 1.5 2.1 Muslim 4925 34.8 33.4 36.1 Others 127 0.9 0.7 1.1 Mult. religion 5 0.0 0.0 0.1 Total 14,218 100.0 High middle income countries No religion 513 3.7 3.2 4.1 Catholic 8736 60.5 59.0 62.0 Protestant 3665 24.4 23.1 25.7 Hindu/Buddhist 66 0.5 0.3 0.7 Muslim 1013 6.9 5.9 7.9 Others 615 3.9 3.4 4.3 Mult. religion 31 0.2 0.1 0.3 Total 14,639 100.0 High-income countries No religion 3488 16.7 15.8 17.6 Catholic 9881 50.4 48.9 51.9 Protestant 4950 24.8 23.5 26.1 Hindu/Buddhist 871 4.3 3.9 4.6 Muslim 103 0.5 0.4 0.7 Others 675 3.1 2.8 3.5 Mult. religion 27 0.1 0.1 0.2 Total 19,995 100.0 All countries No religion 7834 16.1 15.6 16.6 Catholic 21,763 45.2 44.3 46.1 Protestant 10,461 21.1 20.3 21.8 Hindu/Buddhist 1273 2.4 2.2 2.6 Muslim 6041 12.4 11.9 12.9 Others 1417 2.7 2.5 2.9 Mult. religion 63 0.1 0.1 0.2 Total 48,852 100.0 55.3%). However, in Mexico (96.2; 47.9%), Romania (99.7; 48.5%), and Bulgaria (97.9; 7.9%), the gap widens between the vast majority who declare a religion and those who consider religion an important aspect of everyday life. In other countries such as in Italy, Ukraine, Spain, Portugal and Northern Ireland where most of the people declared a religion, only between a third and a quarter of the population declared the religion as a very important in their daily life. In the two Chinese samples these percentages are extremely low: 1.6 and 1.4%. Post-communist countries seem more diverse in their relationship with religion. In the three former Soviet countries (Romania, Bulgaria, and Ukraine), the vast majority of respondents declare a religious affiliation. In Romania, fewer than half of the population (42.7%) reported seeking comfort from religion when faced with problems, in Ukraine this drops to nearly one fourth (23.8%) and in Bulgaria only 6.4% seek comfort from religion. A similar pattern was seen in these three countries when asked if respondents consider what their religious advisors would suggest when making a decision. Only a small percentage of those interviewed report attending religious services more than once a week in these countries 5.9% in Romania, 1.7% in Bulgaria and 1.2% in Ukraine. Less than 10% of respondents in each of the two sites in China declared a religion, and less than 2% said religion is important in their daily life. Overall, 41.2% of participants in LLMI countries are frequent attendants, 11.4% in UMI countries and 6.8% in HI countries (Table 5). Only in Nigeria and Iraq the majority of those who declared a religion attend religious services more than once a week (76.0 and 91.7%, respectively). In five countries (Brazil, South Africa, the USA, Northern Ireland and Colombia) 11.6 18.8% of respondents are frequent attendants of religious services. In the remaining countries this percentage is lower than 10%. As expected, China has the lowest rate of frequent attendants with 0.4% in Shenzhen and 0.9% in Beijing/ Shanghai. and working status (to be a student OR 0.7; 95% CI 0.5 0.9 and a homemaker OR 1.4; 95% CI 1.1 1.7) (available on request). Declaring a religion does not necessarily mean that religion is important for the person in his or her daily life. In some countries where most of the respondents declare a religious affiliation, the majority also finds religion to be important in their everyday life. These results hold for Nigeria (99.8% declare religion; 93.1% say it is important in their daily life), Iraq (100; 80%), Lebanon (100; 66.7%), Colombia (95.4; 68.7%), Brazil (91.9; 64.5%), South Africa (95.9; 63.9%) and to a lesser extent the USA (85.9; Discussion There is limited epidemiological research on use of religious providers for mental health problems with standardized mental health assessments. Comparisons with existing data are rendered difficult due to differences in the period of service use covered (1 year versus lifetime) or the selected population (those who have lifetime disorders, those who are in contact with services for mental health problems). Previous community surveys seem to corroborate our data and findings. We found similar sociodemographic

362 Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 Table 4 Logistic regression predicting ANY use of 12M religious counseling who used any services in past 12M Variables OR OR lo OR hi p value Average for all countries as baseline 1.0 N/A N/A N/A Belgium 1.0 0.3 3.4 0.96 Brazil 0.8 0.5 1.3 0.34 Bulgaria 0.7 0.3 2.0 0.54 China (Benjin, Shanghai) 3.1 1.2 8.2 0.02 Colombia 0.5 0.2 1.0 0.05 France 0.9 0.4 1.8 0.72 Germany 2.3 1.1 4.8 0.02 Iraq 1.0 0.4 2.4 0.98 Italy 1.2 0.7 2.2 0.50 Japan 3.1 1.3 7.4 0.01 Lebanon 1.7 0.8 3.8 0.18 Mexico 0.4 0.2 0.9 0.02 N. Ireland 0.3 0.2 0.6 \0.0001 Netherlands 1.1 0.5 2.4 0.79 Nigeria 1.3 0.5 3.2 0.61 Portugal 0.5 0.3 0.8 0.00 Romania 0.3 0.1 0.7 0.01 Shenzhen 0.9 0.3 2.8 0.92 South Africa 1.9 1.4 2.7 0.0001 Spain 0.3 0.1 0.6 0.00 Ukraine 4.8 2.8 8.0 <0.0001 USA 1.4 1.0 1.9 0.05 Gender male 1.0 N/A N/A N/A Gender female 1.0 0.8 1.3 0.97 Age 18 34 1.0 N/A N/A N/A Age 35 49 0.8 0.6 1.1 0.13 Age 50 64 0.4 0.3 0.6 \0.0001 Age 65 0.4 0.3 0.6 0.0001 Married 1.0 N/A N/A N/A Separated/widowed 1.4 1.1 2.0 0.02 Never married 1.2 0.9 1.7 0.18 Low income 0.9 0.6 1.2 0.44 Avg low income 1.0 N/A N/A N/A Avg high income 1.0 0.8 1.4 0.82 High income 0.9 0.6 1.3 0.51 DX severity: severe 1.1 0.8 1.5 0.59 DX severity: moderate 1.1 0.8 1.4 0.73 DX severity: mild 1.1 0.7 1.7 0.73 DX severity: none 1.0 N/A N/A N/A Religion-none/atheist 1.0 N/A N/A N/A Religion-catholic 1.2 0.5 2.8 0.60 Religion-protestant 1.4 0.6 3.0 0.46 Religion-all others 1.6 0.7 3.8 0.28

Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 363 Table 4 continued Variables OR OR lo OR hi p value How often do you attend religious services? Never* 1.0 N/A N/A N/A \1 time per month 1.3 0.8 2.0 0.32 1 3 times/month 2.8 1.7 4.4 <0.0001 About 1 time/week 3.5 2.2 5.6 <0.0001 [1 time/week 6.5 3.9 10.8 <0.0001 When you have problems or difficulties, how often do you seek comfort through religious or spiritual means? Never* 1.0 N/A N/A N/A Rarely 2.7 1.2 5.9 0.02 Sometimes 2.4 1.1 5.5 0.03 Often 3.6 1.5 8.8 0.0042 In general. how important are religious or spiritual beliefs in your daily life? Not at all important* 1.0 N/A N/A N/A Not very important 0.9 0.4 2.1 0.74 Somewhat important 0.9 0.3 2.2 0.78 Very important 1.1 0.4 2.9 0.82 When you have decisions to make in your daily life, how often do you think about what your religious or spiritual beliefs suggest you should do? Never* 1.0 N/A N/A N/A Rarely 0.5 0.3 0.9 0.02 Sometimes 0.7 0.4 1.2 0.17 Often 0.8 0.4 1.5 0.51 Modified by income group only OR 0.2 (0.1 1) p = 0.05; OR 2.3 (1.1 4.7) p = 0.03 Bold for country who are above 1 and italic for those below; bold italic for the reference * Missing responses and responses 8 ( don t know ) and 9 ( refused ) were added to this category variables that were associated with the use of religious providers than those found in previous studies [16, 17]. Our data are also comparable to those found in the National Survey on Drug Use and Health (NSDUH). Chinese US-born or immigrants were much less religiously affiliated than the general US population: 50% declared a religious affiliation; 0.6% of Chinese immigrants and 4.6% of US-born Chinese reported a 12 month use of religious advisors for mental health problems as compared to 0.3% in our Chinese sample (0.1% in Shenzhen) showing a sort of gradient toward religious affiliation and religious advisors use from China mainland to the US born Chinese [18]. The relative importance of religious providers in Africa is in line with a previous community survey from Nigeria, reporting that the preferred treatment options for mental disorders were religious healers, traditional healers, and use of formal health care services with 41, 30 and 29%, respectively [19]. Similarly, a study conducted in Singapore [20] examined the use of religious advisors in a large representative sample of the population: only 0.3% reported seeking help from religious or spiritual advisors for help with psychological problems during the last 12 months. This rate is similar to rates reported in Asian countries from the WMH surveys: Japan 0.6%, China 0.3% and Shenzhen 0.1%. These countries were also the least religious countries in our sample and this was reflected by their low use of religious advisors for mental health problems. In LLMI countries specifically, the use of religious advisors was higher in respondents with the most severe disorders. This pattern was not found in UMI or HI countries. However, while this trend was not present in these income groups an important percentage of those suffering from the most severe forms of disorders was in contact with religious providers with as much as 21.6% in South Africa and 16.4% in the USA which is corroborated by studies which reported that clergy is in contact with persons who suffer from severe mental health disorders, alcohol addiction, and severe depression with suicide risk [4]. Our data point to the importance of religious providers in countries with low resources for professional mental health infrastructures. This is a particularly important

364 Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 Table 5 Prevalence estimates for religion practice variables by country, country income-level and all countries Declaring a religious preference You attend religious services MORE THAN ONCE PER WEEK Religious/spiritual beliefs are VERY IMPORTANT in your daily life You OFTEN seek comfort through religious/spiritual means when experiencing problems in family, work, or personal life You OFTEN think about what your religious/ spiritual suggest you should do when you make decisions in your daily life Country Percent S.E. Freq Percent S.E. Freq Percent S.E. Freq Percent S.E. Freq China* 7.5 0.7 0.3 1622 1.6 0.4 1622 1.9 0.4 1622 2.0 0.6 1622 Colombia 95.4 11.7 1.1 2373 68.7 1.5 2373 48.0 1.7 2373 48.1 1.6 2373 Iraq 100 91.7 0.7 4277 80.0 1.1 4277 61.6 1.4 4277 62.6 1.3 4277 Nigeria 99.8 76.0 1.2 2133 93.1 0.8 2133 77.9 1.3 2133 76.5 1.2 2133 Shenzhen* 9.8 0.4 0.2 2361 1.4 0.4 2361 0.9 0.2 2361 0.7 0.2 2361 Ukraine 83.9 1.2 0.4 1452 22.2 1.8 1452 23.8 2.1 1452 24.9 1.9 1452 Low low middle 71.9 41.2 0.7 14,218 52.1 0.7 14,218 41.0 0.6 14,218 41.2 0.6 14,218 Income Brazil 91.9 16.9 1.0 2915 64.5 1.2 2915 54.1 1.6 2915 48.6 1.6 2915 Bulgaria 97.9 1.7 0.4 2111 7.9 0.6 2111 6.4 0.7 2111 5.2 0.5 2111 Lebanon 100 8.6 1.2 1027 66.7 1.8 1027 56.6 2.1 1027 52.2 2.7 1027 Mexico 96.2 8.4 0.9 2354 47.9 1.9 2354 25.6 1.5 2354 23.0 1.4 2354 Romania 99.7 5.3 0.5 2356 48.5 1.2 2356 42.7 1.4 2356 42.3 1.4 2356 S. Africa 95.9 18.8 1.1 3876 63.9 1.1 3876 42.6 1.2 3876 35.6 1.1 3876 Middle 96.3 11.4 0.4 14,639 51.0 0.6 14,639 37.9 0.6 14,639 34.0 0.6 14,639 high-income Belgium 79.6 1.2 0.5 1022 11.2 2.2 1022 10.0 1.5 1022 6.2 1.2 1022 France 80.1 1.0 0.3 1436 9.6 1.2 1436 8.0 1.0 1436 7.2 1.1 1436 Germany 77.7 2.3 0.5 1320 8.8 1.2 1320 7.4 1.2 1320 6.4 0.9 1320 Italy 96.8 6.2 0.7 1779 23.4 1.4 1779 22.5 1.5 1779 16.0 1.3 1779 Japan 57.5 2.0 0.6 1536 9.3 1.0 1536 2.8 0.5 1536 2.1 0.4 1536 Netherlands 60.5 2.9 0.6 1092 14.6 1.1 1092 11.4 1.3 1092 10.0 1.2 1092 N. Ireland 95.2 11.6 0.9 1962 31.1 1.5 1962 28.1 1.2 1962 23.5 1.2 1962 Portugal 88.5 4.5 0.5 2058 32.3 1.5 2058 27.4 1.2 2058 22.2 1.2 2058 Spain 86.9 4.5 0.8 2119 21.3 1.5 2119 13.8 1.4 2119 13.6 1.4 2119 USA 85.9 12.5 0.9 5671 55.3 1.4 5671 45.5 1.3 5671 40.3 1.1 5671 High-income 83.3 6.8 0.3 19,995 29.8 0.5 19,995 24.4 0.5 19,995 20.9 0.4 19,995 All countries 83.9 18.2 0.3 48,852 42.6 0.3 48,852 33.3 0.3 48,852 30.7 0.3 48,852 *Very low endorsement rates of religiosity: the percentages about religious paractice in these countries should be interpreted with caution because they are based on small numbers of cases consideration for countries categorized as low to lower middle income and thus, where formal resources are largely non-existent. For instance, in LMI countries only about 1.9% and in low-income countries only 0.5% of the health budget is allocated to mental health. This means that on average there is a scarcity of both outpatient and inpatient treatment facilities [21] and a substantial literature supports the important role of the informal sector including religious advisors in countries that lack formal medical resources [22 24]. In addition, religious providers provide counseling to people with serious mental health problems [3, 25, 26]. This is particularly problematic because of less than half of the clergy received any formal training in clinical care, they did not feel confident in their ability to deal with mental health issues, or they did not have sufficient contacts with formal MH system [4, 6]. This suggests an important need for professional development training activities around mental health that many religious advisors would welcome since these clergy viewed depression as a common and serious problem but viewed time constraints and lack of formal training in counseling and referral as barriers to providing care [27]. In addition to availability of resources, country s religious values also likely play a role in the prominence of and preferences for seeking help from religious advisors. In most of the predominantly religious countries (in which more than 50% of the population declared that religion is very important in their daily life), the reliance on religious

Soc Psychiatry Psychiatr Epidemiol (2017) 52:353 367 365 advisors who access services for mental health problems was higher than average. But there are some additional structural factors of mental health care organization as well as variation in the perception of religion or clergy. For example in a previous paper of several European countries, we had suggested that one possible explanation for the surprising low rate of religious providers use in Spain in contrast with rates found in Italy may be due to the role played by the clergy in the Spanish civil war [10]. Indeed in highly traditional religious countries such as Portugal, Northern Ireland, Romania, Colombia and Mexico, once controlled for religiosity, the trend of use of religious providers is surprisingly low as compared to the other countries as in Iraq in the LMI regression model. Indeed considering that, among reasons evoked for visiting a religious advisor versus a MH provider, a high stigmatisation of mental health problems has been evoked, negative attitudes of the religious providers in some religions and cultures especially in Arabic cultures toward MH may refrain people for addressing to them for these problems [7]. On the contrary, living in the Ukraine, South Africa, USA, Germany or Japan increased the use of religious providers after controlling for demographic, clinical and religiosity factors. High visibility of clergy in these countries may attract people especially when access to mental health specialists is difficult to obtain or inexistent. China was surprisingly in this last group although religious affiliation is rare and difficult to maintain. Our results should be interpreted in light of several limitations. First, response rates varied across WMH surveys, but this was controlled for through post stratification adjustments. Second, some clinically important disorders such as schizophrenia were not assessed in WMH surveys because earlier validation studies have shown that they are overestimated in lay-administered [11]. However, these studies have also shown disorders such as non-affective psychosis usually are comorbid with anxiety, mood, or substance disorders. Therefore, they would be captured in our analyses which used severity levels. A third potential limitation is that the reliability and validity of diagnoses made with the WMH CIDI might vary across countries. Although acceptable concordance has been noted between diagnoses made with the CIDI and those from blinded clinical interviews, such studies have been done almost exclusively in developed countries [28]. The accuracy of CIDI diagnoses could be lower in other countries. One distinct possibility is that there is a lower relevance of CIDI symptom descriptions in low and low middle income countries than in high middle and high income countries, or greater reluctance to endorse emotional problems in countries with short traditions of free speech and anonymous public-opinion surveying. Fourth, without corroborating data for service use we could not examine the accuracy of self-reported treatment use or how validity of self-report differs across specific sectors, clinical or sociodemographic characteristics, or cultural groups. Despite the unprecedented scope and size of the WMH survey initiative, some analyses included small numbers of respondents, which might have limited our interpretation of the findings. In addition, the investigation of religion carries inherent problems due to its sensitive nature in certain political and cultural contexts. In some countries not having a religion is a crime, rendering the report of no religion impossible even though anonymity was guaranteed. Conversely, in other countries reporting of certain religious affiliations may be considered unacceptable. Pooling Protestants together concealed heterogeneity in the protestant category and dichotomising religious attendance into two categories are also limitations. Limitations Entry question on religion is focused on organized religion and excludes spirituality outside religion. Mental health questions derived from a unique questionnaire: the CIDI, carefully translated but this does not preclude cultural dissensions in recognition and acceptance of the diverse mental health symptoms. Protestants have been pooled together in the analyses when there is a large heterogeneity in the protestant category probably greater than the differences between Protestants and Catholics. Conclusions In distress, many people turn to trusted individuals in their social networks, including religious advisors [5]. Thus, many religious advisors function as first responders for individuals facing mental health challenges. Our data enable us to conclude that religious providers play an important role in the broad MH field. They may even be the central entry point to further care so their training is essential in mental health resource planning. This is especially important in LLMI countries where those with the most severe disorders had the highest rates of contact with religious advisors for mental health problems. They can also represent an important and more available resource in some low income settings, where access to traditional forms of mental health care are impeded by the scarcity of trained professionals and financial resources and so could represent potential for scaling up. Training of religious advisors regarding mental disorders and interventions could be modeled on the WHO mental health Intervention